Conceptual links and a nascent literature suggest that in-home caregiving arrangements have a bearing on disabled older adults' use of medical services. Larger caregiving networks, the presence of unpaid help, lower levels of caregiver burden and higher levels of reward, and the absence of unmet need with activities of daily living have previously been linked to desirable health outcomes and a lower likelihood of permanent nursing home entry for recipients. Here we extend this line of inquiry by hypothesizing that these attributes of in-home caregiving arrangements are indicative of the depth of available resources that may be mobilized to assist recipients with personal care as well as necessary medical care. The proposed project examines the relationship between these attributes of in-home care arrangements and hospital outcomes that are thought to be particularly sensitive to the ability of in-home caregivers to facilitate appropriate medical care. This prospective cohort study draws on data from the 1999 National Long Term Care Survey, its accompanying Caregiver Survey, and Medicare claims data for the 12-month period following completion of the 1999 survey. The relationship between in-home caregiving arrangements and hospitalization outcomes (specifically, potentially preventable hospitalizations, repeat hospital admissions, and delayed discharges) will be estimated using multivariate regression models that control for individual and caregiver characteristics known to be related to the outcomes of interest, with particular attention to patient morbidity and severity of illness. The proposed research will provide an understanding of the influence of in-home caregivers in an important area that is beyond the scope of outcomes typically attributed to them. The ultimate goal of this work is to develop strategies that support caregivers and improve the health of older adults. Findings from this pilot project will inform measurement and data collection efforts, as well as interpretation of results from a multi-component caregiver intervention in Baltimore, Maryland that is expected to begin in July 2006. [unreadable] [unreadable] [unreadable]